Author :
Verdulla, F.M. ; López, M.J. ; Prian, M.
Abstract :
In this paper, it is proposed a method for controlling chaotic systems; where the main goal is to obtain a periodic behaviour for a chaotic system. In our approach, the system to control is considered as a black-box, and therefore it is not necessary to know a mathematical model of the system, only experimental measurements are used. Our method employs pulses with adjustable amplitude and width, and it is implemented in discrete time. In order to generate pulses control, a variable Poincare section is used; which is computed online using a moving average sampling signal. Measurement noise is considered too, by means of an additional controller parameter (hold-off time), resulting that controller tuning is made using four parameters: proportional gain, sampling time, pulse width and hold-off time. In order to test the proposed method, computer simulations with several representative chaotic systems (Lorenz, Chua, Chen, Colpitts and others) are carried out and satisfactory results are obtained.
Keywords :
Poincare mapping; chaos; control system synthesis; discrete time systems; moving average processes; nonlinear control systems; nonlinear dynamical systems; periodic control; signal sampling; Chen system; Chua system; Colpitts system; Lorenz system; black-box system; chaotic system control; controller parameter tuning; discrete time system; hold-off time; mathematical model; measurement noise; moving average sampling signal; nonlinear dynamical system; periodic behaviour; proportional gain; pulse width; pulsed control method; sampling time; variable Poincare section; Chaos; Control system synthesis; Control systems; Mathematical model; Proportional control; Pulse generation; Pulse measurements; Sampling methods; Signal generators; Space vector pulse width modulation; Control system; chaos control; chaotic system; nonlinear dynamics; periodic orbit; pulsed control;
